Acute myeloid leukemia (AML) is a fatal disease in which most patients die despite achieving initial complete remission (CR). Even under very aggressive multi-agent chemotherapy regimens and myeloablative allogeneic stem cell transplantation, relapse rates are high. Thus, understanding the mechanisms that lead to relapse is critical. Increasing evidence suggests that AML is originated and maintained by a subpopulation of leukemic stem cells (LSCs), which are resistant to standard chemotherapy and thereby provide a reservoir of cells that drive disease relapse. LSCs reside in a unique physiologic state from their normal counterparts and have acquired addiction to pathways such as NFkappaB for survival. We have recently discovered that LSCs demonstrate increased autophagy and thus hypothesize that autophagy is a mediator of LSC survival and chemoresistance. We propose the following specific aims: (1) to test the hypothesis that autophagy mechanisms are upregulated in LSCs when compared to their normal counterparts, conferring survival advantages and chemoresistance properties on LSCs; (2) to test the hypothesis chemical or genetic perturbation of the autophagy pathways will decrease LSC survival and chemoresistance; (3) to determine the ability of known anti-LSC drugs to inhibit the autophagy process. The evaluation of autophagy as a mediator of LSC chemoresistance and relapse is critical towards better delineating the unique features of LSCs that can be capitalized upon towards improving AML therapy. This proposal will address whether autophagy represents a feasible therapeutic target and/or sensitizes LSCs to induction therapy, diminishing the likelihood of relapse.